Harvesting Volatility Risk Premium

Thanks to Bankeronwheels, last week I read this post from Early Retirement Now on trading options.

A quick self-reflection digression. I know the existence of ERN because Mr.Rip once told me about it. I had a peek at the blog and immediately my brain rejected it because of “bad layout – too much text”. This is a funny thought for someone who runs a blog with a very bad layout. Years ago I wrote to the guys behind Monevator asking…a general opinion on TILS (obv it was just a lame attempt to make them aware the blog existed). They kindly came back to me with constructive comments like “the layout sucks”. To which I reacted in the very mature manner of said the guy who runs the 90s Amazon website in 2018. As you can see, the crappy layout is still here despite I know how important is to use a nice one. Welcome to my world.

Anyway, back to ERN option primer. Despite my initial reaction, I read some posts throughout the years, the latest one about that 100% stocks in retirement paper, and I always found them well-researched and written. That’s why this post was something like a shocker to me. You should really read it…but I know you won’t so, here is a very brief summary. He (is he a he? let’s assume so) sells short-dated, far out-of-the-money naked options on margin for additional income. Yep, like those cheap influencers you find on YT. Here is a summary of the option strategy returns:

An Information Ratio of 3, FOCKING HELL (I just watched Bullet Train and I looove Aaron-Taylor Johnson accent).

Look, you cannot put forward a strategy as “Hey, it is a great way to complement your other passive income sources” and “I do not want to be 100% transparent about the strategy”. Can an IR of 3 1 be achieved by trading vol? Yes, Kris Sidial and Cem Karsan do similar stuff but they definitely do not go out sort of winking “it is easy, you can do it too”. You see my issue here? It is not the implausibility of the returns, it is them combined with the anyone can do it message that threw me out of my chair.

That plus some bs sprinkled on top like “this strategy has a conservatively 0.5 correlation to stocks”. Bruv, you are selling options on SPX backed by a stock portfolio, how do you think they will both behave when SPX is down 4% in a day?!? Cause that’s the only correlation that matters, innit?

How do I go about it? If you asked me before I read this post, I would have 100% bet that ERN was way smarter than me. He brings some math/stat on the table that I am not even able to argue with. I will try to lay out the limited data that I have and see where it lands.

First Principles?

I can replicate a short put position with a portfolio of borrowings plus some long on the underlying (I’ll save you the Black-Scholes formula). This portfolio should generate a return, also in risk-adjusted terms, that is similar to plain vanilla Buy&Hold because…same ingredients? Ultimately, you are exposed to the same risks. There are some products that confirm this, for example $PUTW, the put writing ETF: it sells two puts per month on the S&P500 backed by a cash collateral and targets better risk-adjusted returns, albeit with lower absolute returns.

PUTW started to trade in 2016 so its performance is affected by the short backtest period.

As it often happens with investment products, the strategy backtest was more promising:

CBOE maintains indexes on both BuyWrite (covered calls) and PutWrite that all points at the same results:

Depending on the period, some strategies perform better than others: there are several parameters than can be changed, the ATM vs OTM strike price or the maturity length. Usually, when a certain configuration seems to prevail, that’s when it stops working (relatively speaking). Markets adapt.

That’s why I find it hard to believe you can go from a Sharpe of 0.3 to 3.

Volatility Premium

“Unless you have a very special edge, there is only one way how you can systematically make money in investing: You need to bear volatility. You need to be exposed to it. You need to suffer when it rises. In finance lingo: You need to short it.” From Fallacy Alarm. Or, as Corey Hoffstein says, risk can only be transformed: if you hedge all your risks, you can only expect to make the risk-free return.

Let’s have a look at some of the purest forms to express a short volatility position, aka volatility risk premium harvesting. This section is here to prove that:

ERN is right when he says that selling puts should generate a +EV
you can harvest the premium only if you stomach high volatility

The $SVXY ETF: ProShares Short VIX Short-Term Futures ETF seeks daily investment results, before fees and expenses, that correspond to one-half the inverse (-0.5x) of the daily performance of the S&P 500 VIX Short-Term Futures Index.

This very naïve strategy has a respectable Sharpe ratio. Conveniently, that deceptive 0.63 market correlation is quite different when it matters (here I show you only the period post-Volmageddon otherwise the graph would unreadable):

See how good I am in drawing circles with a mouse?

Let’s now look at a more advanced strategy, $SVOL ETF: The Simplify Volatility Premium ETF (SVOL) seeks to provide investment results, before fees and expenses, that correspond to approximately one-fifth to three-tenths (-0.2x to -0.3x) the inverse of the performance of the Cboe Volatility Index (VIX) short-term futures index while also seeking to mitigate extreme volatility. The fund’s short VIX position provides investors an optimized exposure for monetizing the premium in the VIX futures market. A modest option overlay budget is then deployed into VIX call options to help protect against adverse moves in VIX.

$SVOL tries to limit the strategy volatility by:

targeting a lower level overall
spending part of the premium to buy protection (in the form of call options)

$SVXY just hold cash as collateral for the short futures while $SVOL holds 23% in two other Simplify bond ETFs to enhance returns.

$SVOL was launched recently so there is not that much to compare to…until the next violent bear market

Actually, there is this paper that created a synthetic backtest for $SVOL:

the above does not include trading costs.

As you can see in the last column, SVOL should (and will, most likely) have a decent Sharpe Ratio accompanied by a crazy Max DrawDown. VRP, the first column, represent the author’s strategy that flips between short and long volatility. A few quick insights from the paper:

VRP is not a complex strategy but is not the most simple either (I think you cannot even get the VIX Future prices for free but I might be wrong)
VRP reduces the drawdown in a meaningful way but there is a negligible improvement in terms of SR

In you look at below figure, you realise the model almost spend the same amount of time long (VIXY) as it does short (SVOL) volatility:

Meaning obtaining a high SR going only short vol, and doing so in an almost systematic fashion, is unrealistic.

Last, the author’s comment on his CTA choice as the “companion” of the vol strategy: The reason behind selecting only commodities and bonds as asset classes is to minimize the correlation against my volatility strategy, which has an equity bias as well.

ERN strategy is just levered stock beta

We saw 3 different takes on vol premium harvesting and all concur on the fact that you should not use stocks as collateral. My humble take is that, if ERN’s strategy is not pure alpha, is simply staking stock beta on top of stock beta. Why his regression analysis does not show it? Simply because the strategy is, so far, miles better than a B&H on SPY or PUT. Sometimes the easy answer is the right answer: he run a long-stock strategy in a period where stocks went up.

It can be pure alpha, and I hope for him it is! But if so, I do not see the point (other than bragging) to share it, since on one else would be able to replicate that performance.

0DTE

ERN talks extensively about the difference in using long-dated options and 0DTE/1DTE…at least from his point of view. All the instruments I presented above do not use 0DTE and 0DTE demonstrated to be quite peculiar:

See how the 0DTE premium actually jumped UP in 2020 compared to all other maturities premium? This confirms ERN’s empirical evidence: there was plenty of juice to extract from 0DTE and the juice has been flowing plentifully. It is a market dominated by vol buyers: 0DTE are perceived as great “lottery tickets” (therefore the juice if you sit on the other side of the table). It is a complex market and few (any?) have a great understanding of what it is going on.

The key words for me are “SO FAR”. Sure, liquidity providers should receive a reward for their activity. Should the reward be so high and, crucially, volatility-less? Probably not. Not that I have anything against ERN but things do not work this way.

“I’d also never recommend using excess leverage. For example, if the strategy did so well with 7.7% return and 2.5% risk, why not run this with an additional 10x leverage and make 77% returns with 25% risk? What can possibly go wrong? Check my post on the “optionsellers” debacle again!” ERN himself wrote this paragraph. If you lever the strategy 4x, targeting a still conservative 10% risk, you generate a 31%, almost linear, annual return.

In 25 years, you go from 10k to 10m. ERN is fine with running the strategy in a very conservative manner but someone else for sure is not. This is why fat pitches like this one cannot be permanent, others will be fine with a lower IR and this will drive down margins for everybody.

This and the fact that you cannot run stocks as collateral, further reducing the whole deal attractiveness.

Risk Management

“Very occasionally, you suffer a loss if options go “in the money” to a stop loss is triggered before then.” The issue with stop loss orders is that you need someone else to be on the other side of your trade. Our Master Cem Karsan teaches us that the only way to hedge a 0DTE is with…a 0DTE. It is unlikely, but not impossible, that one day the whole market will sit on the same side, trying to get out at the same moment. At that point, the market will simply spiral down, jumping prices and most likely your stop loss order. Everything is liquid and continuous and balanced until it isn’t.

4% declines in one day are rare but not like 1-in-200 years rare. You have to pray that the day before you realised the risk and moved the OTM strike level from 2%/3% way lower.

“You don’t go from a 2006-style stock market to the 2008 Lehman Brothers failure literally overnight.” This is true but ERN is looking at the wrong risk. Think about how many heard about Volmageddon. That was exactly a one day event with rather mild consequences outside that niche. Plus, if your strategy is not correlated to stocks…why do you refer to stock market failures? I cannot tell you when, I cannot tell you how, but I am SURE the “0DTE party” will manifest in a way (almost) no one envisioned. Like…a Black Swan event.

When the music stops

Considering the high transaction costs for retail investors, it is also possible that the strategy will simply stop being profitable by a gradual squeeze in premiums. ERN is already experiencing this, mentioning he went from targeting a premium of $0.5 to $0.1/$0.2; he also reacted the predictable (greedy?) way: At the market open, if my puts that expire later that day appear “safe,” I am comfortable selling additional 0DTE puts expiring that day as well.

The squeeze can happen for 2 reasons:

more investors are lured in by the fat margins (and low perceived risk)
less investors are attracted by lottery tickets, the long 0DTE side

Conclusion

I do not want to “blame” ERN for finding an opportunity and riding it. That’s what investing is all about.

The 0DTE market is still in its infancy. I have heard many smart and experienced investors scratching their heads about it, having more questions than answers. Framing it as “passive income” feels pretty reckless. A six-year backtest is not long enough for something that might have a very fat tail and dismissing the correlation with stocks is a big mistake.

That said, farming the RobinHood YOLOers that gamble on long 0DTEs might as well continue profitably for a long time. Short vol is absolutely the trade. But I would put it in a portfolio INSTEAD of stocks, not on top of it. Like 55% SPY and 5% SVOL instead of 60% SPY. Or a short 0DTE strategy with a more conservative collateral (I do not think there is a convex hedge here; $TAIL, $CAOS, $BTAL might not even move the day of the 0DTE collapse. Maybe spending some premium on a deep OTM call but at that point trading costs might kill you). Fact is, the moment you de-risk your collateral, you might lose more in terms of expected returns than what you get with the option overlay.

IMHO the savviest way to see this is that you had a lucky streak at the casino, you won the lottery ticket, and it is better to exit while you are ahead than risking to giving it all back. And maybe more.

What I am reading now: